Benchmarking the effect of weather data upon …...CIBSE Technical Symposium, Liverpool John Moores...

CIBSE Technical Symposium, Liverpool John Moores University, Liverpool, UK, 11-12 April 2013

of 12

Benchmarking the effect of weather data upon energy estimation of UK homes

Gavin Murphy, John Allison, John Counsell

Department of Mechanical and Aerospace Engineering, University of Strathclyde, Glasgow.

[email protected] Abstract This paper extends a fundamental dynamic method of assessing the controllability of a building and its servicing systems: IDEAS – Inverse Dynamics based Energy Assessment and Simulation. IDEAS produces results which are calibrated with the UK Government’s Standard Assessment Procedure (SAP). The extension presented is a measure of the effect upon energy estimation by varying climactic data, using CIBSE TRY/DSY weather data for 14 locations across the UK. A calibrated standard test case dwelling is initially modelled in IDEAS and SAP. Using each of the CIBSE weather locations, the variation in energy estimation of the standard test case dwelling is analysed. Results suggest that use of localised weather data can have a noteworthy effect on energy estimation and can play an important role on delivering buildings which are truly fit for purpose. Keywords Dwellings, Standard Assessment Procedure (SAP), IDEAS, CIBSE TRY/DSY weather data 1.0 Introduction The European Directive on the Energy Performance of Buildings [1, 2], referred to as the Energy Performance of Buildings Directive (EPBD) stipulates that all European member states must produce an Energy Performance Certificate (EPC) whenever buildings are constructed, sold or rented. In the UK, SAP is the procedure used to generate an EPC. SAP is the recognised method for building professionals to meet buildings compliance and is the culmination of three decades of research commencing with BREDEM 1 [3, 4].

Figure 1 - Sample SAP derived Energy Efficiency and Environmental Impact Ratings for Scotland

Studies have shown that there can be variances in results between SAP and dynamic simulation tools [5]. Fundamental dynamic methods have been shown to be


of 12

relevant for controllability analysis [6-8] and for the assessment of buildings [9]. A benefit of dynamic systems modelling is that a state space model [10] can rapidly and thoroughly determine the effect of disturbances such as free heats gains or external temperature [11]. A novel advanced dynamic calculation method (IDEAS) of assessing the controllability of a building and its servicing systems has been developed. IDEAS has been calibrated over a large range of test cases with SAP [12]. SAP estimates the energy estimation of a dwelling based upon a notional central location for the UK, taken as East Pennines region, this allows for dwellings to be compared on a like-for-like-basis as one weather location allows for one set of solar and one set of external temperatures to be used for each dwelling evaluated in SAP. External temperatures and solar gain can have a critical impact on the energy estimation of a home. In addition to this, factors such as occupancy and appliance heat gains are assumed based upon the total floor area of the dwelling. The assumptions made in SAP, how they compare with BREDEM, and SAP’s use for EPBD compliance has been previously researched and published [13]. Given SAP’s role in energy estimation and its use of specific assumptions for all dwellings, it is important to quantify the effect of using the current single climactic location in comparison with the use of several more localised climactic locations. The focus of this paper is measuring the effect upon energy estimation by varying external temperature data, using CIBSE TRY/DSY weather data for 14 locations across the UK. 2.0 METHODOLOGY

2.1 Developing a modelling environment

An IDEAS based modelling environment was selected for this study. IDEAS allows for an extension of SAP in many areas, such as: the ability to make use of various dynamic weather files [14] and the flexibility to amend the heating setpoint which is tracked (for example, comparisons can be made between tracking a constant setpoint vs. a varying setpoint). IDEAS has been described in depth [15, 16]. Any set-point can be tracked in the IDEAS method, the set-point tracked in this IDEAS model is based upon the standard SAP temperature demand profile (see Figure 2).

Figure 2 - Standard SAP temperature demand profile (a) weekdays; (b) weekends


of 12

By tracking the standard SAP temperature demand profile setpoint, IDEAS will calculate what the predicted energy consumption for that modelled dwelling over a year. This energy use will be affected by factors such as weather data and heat gains from appliances. In IDEAS, heat gains such as external weather data and heat gains from appliances are described as disturbances to the model. These disturbances are taken into account by IDEAS and play an important part in the overall energy estimation analysis of a dwelling. For example, if there are more heat gains then less heating could be required to be produced by the heating system for the standard SAP temperature demand profile to be met. The focus of this paper is the effect of the external temperature component from CIBSE TRY/DSY weather data for 14 locations across the UK and measuring the impact of this weather variation on the predicted energy performance of dwellings. 2.2 Disturbance heat gains In this IDEAS model dynamic time-varying solar gains and external temperature are used. A disturbance heat gain into the building is used that is a combination of solar radiation, metabolic heat gains, and heat gains from electrical devices. Disturbance free heats gains must also be calibrated, which is comprised of weather profiles, lighting, appliance and metabolic gains. SAP makes assumptions for factors such as occupancy and hot water use based upon the ‘total floor area’ of a dwelling; a single representative weather location (taken as East Pennines, UK) is used as the basis for solar gains. Appliance gains were taken from an International Energy Agency / Energy Conservation in Buildings and Community Systems Program (ECBCS) Annex 42 study based upon real UK test data for 69 monitored buildings [17]. Appliance gains were fixed for each of the test cases in this study. Solar radiation data for the East Pennines, UK was imported into the IDEAS model for calibration with SAP, using a data file from Meteonorm [18]; this was used to provide data for the solar radiation. Solar data was fixed for each of the test cases in this study. CIBSE TRY/DSY weather data for 14 locations across the UK was used to provide yearly external temperature for each of the test cases. The modification of the external temperature and subsequent analysis on the energy performance of dwellings is the focus of this study.

2.3 Baseline Calibration results

This IDEAS model has been calibrated against SAP over a range of test cases; in this paper one test case is presented as a baseline: Standard Test Case (unfilled cavity construction). A standard construction dwelling was modelled, with a structure U-Value of 1W/m2K, which is representative of unfilled cavity wall structure. Full test case parameters are provided in Appendix A, Table A1.


of 12

The following results were obtained when taking the mean monthly temperature and energy values from SAP and comparing these with the calculated mean monthly temperature and energy values calculated by this IDEAS model.

Figure 3 - Test Case 1: mean monthly zone temperature comparison

Figure 4 - Test Case 1: mean monthly energy comparison

Figure 3 and Figure 4 demonstrate the similarity between IDEAS and SAP outputs when monthly values are plotted against each other. From the data used to plot Figure 3, monthly zone temperatures, a data match of 98.9% was achieved when the arithmetic means (AM) calculated from SAP (AM = 19.77°C) and IDEAS (AM = 19.99°C) outputs were compared. Similarly for Figure 4, monthly energy consumption a match of 96.16% was found between an IDEAS AM of 17,601kWh/year and a SAP AM of 18,304kWh/year. Based upon results from the standard test case, modelling a dwelling of unfilled cavity construction, there is close match between this IDEAS model and SAP for both temperature and energy consumption. The standard test case and calibration highlighted in Figures 3 and 4 are based upon an external temperature profile for Manchester, as taken from the CIBSE weather data. The standard test case is used a baseline for the remainder of this paper analysing the effect of varying external temperature on energy estimation.


of 12

3.0 CIBSE TRY/DSY weather analysis 3.1 Introduction CIBSE Test Reference Years (TRYs) and Design Summer Years (DSYs) are available for 14 locations across the UK. The importance of weather data has been highlighted: “weather data has now become an essential component of virtually every new building design and major refurbishment”[19]. Table 1 provides the results from a statistical analysis performed on the CIBSE weather data.

Table 1 - Weather data statistical analysis

Location Latitude (°N) Mean Temperature (°C)

Standard Deviation

Coefficient of Variation

Glasgow 55.97 8.6685 5.4258 0.625921

Edinburgh 55.95 8.8041 5.1438 0.584251

Newcastle 54.9833 9.5952 4.9305 0.513851

Belfast 54.6 9.1794 4.9679 0.541201

Leeds 53.81 10.1719 5.5657 0.547164

Manchester 53.48 10.0023 5.2695 0.526829

Nottingham 52.97 9.6261 5.6518 0.587133

Norwich 52.65 10.1009 5.4631 0.540853

Birmingham 52.48 9.8879 5.8724 0.593898

Swindon 51.5642 9.8772 5.535 0.560381

London 51.5171 11.4412 5.8185 0.508557

Cardiff 51.478 10.4426 4.8871 0.467996

Southampton 50.9339 10.9566 5.4454 0.496997

Plymouth 50.3706 11.1022 4.3692 0.393544

The locations are ordered by their latitude from the top of the British Isles to the bottom. The latitude given is for the city, which is an approximate location of actual CIBSE weather location. The standard deviation is provided to highlight the variability of the temperature of each location from its mean temperature over the year. Results indicate a reasonably large temperature swing from their mean, in line with the wide variance expected of British weather. The coefficient of variation (CV) is an indication of how the standard deviation relates to the mean. The closer the CV is to zero, the greater the uniformity of the weather. As highlighted above, the weather is generally more varied in Scotland and the North of England as opposed to Wales and Southern England.


of 12

Figure 5 – Minimum and maximum external temperatures for each CIBSE TRY/DSY Weather Location

Figure 5 highlights the minimum and maximum external temperatures for each CIBSE TRY/DSY weather location. It can be viewed from Figure 5 that there is a variance of approximately 8ºC between the maximum temperatures and also the minimum temperatures when comparing the weather locations as a whole. 3.2 Distribution Spread

Figure 6 – CIBSE Weather location distribution spread

Figure 6 highlights the distribution spread of the CIBSE TRY/DSY weather files. The markers placed on figure 6 correspond with the locations of the available weather data. It is clear that although the main population density areas of the UK are covered, there are large areas where there is no weather data available.


of 12

4.0 RESULTS 4.1 Yearly calculated dwelling energy use for each CIBSE weather location Table 2 presents the results of 14 IDEAS simulations based upon the Test Case 1 dwelling. Table 2 presents each of the CIBSE weather locations sorted by the yearly energy required for the standard test case dwelling (figures 3 and 4) to meet the Standard SAP temperature demand profile (figure 2).

Table 2 – Results highlighting yearly energy use for Test Case 1 for each CIBSE weather location

Number Location Yearly Energy

(kWh) Variation from

(8), kWh Variation

from (8), %

1 Glasgow 20199 2598 14.76% 2 Edinburgh 19918 2317 13.16% 3 Belfast 19097 1496 8.50% 4 Nottingham 18551 950 5.40% 5 Newcastle 18376 775 4.40% 6 Birmingham 17913 312 1.77% 7 Swindon 17885 284 1.61% 8 Manchester 17601 0 0.00% 9 Leeds 17422 -179 1.02%

10 Norwich 17411 -190 1.08% 11 Cardiff 16519 -1082 6.15% 12 Southampton 15725 -1876 10.66% 13 Plymouth 15117 -2484 14.11% 14 London 14985 -2616 14.86%

Weather location 8, Manchester, is taken as the weather location most representative of that used in SAP; the variation of each weather location from the Manchester data analysis is presented. It was found that the weather locations where most energy is required is Glasgow and Edinburgh. The location where the least energy is required is London. It might have been expected that Southampton, Plymouth and Cardiff would require less yearly energy due to their southerly location. However, London’s mean temperature is the highest out of all of the CIBSE weather locations (see Table 1) and this will have a major bearing of the predicted energy consumption. 4.2 Implications for SAP and EPC generation The current method of SAP, SAP 2009 [20] generates EPCs for dwellings in the UK based upon a notional centre of the UK – the East Pennines location. This is taken as a representative location for the UK as a whole. Mean external temperatures are derived from this location. From Table 2, location number 8, Manchester is the most applicable CIBSE weather location for the SAP East Pennines region. The results for Manchester can be seen to be centrally distributed between the other weather locations: this is to be expected based upon the approximately central location of Manchester. Table 2 highlights that there is a 15% difference between results when the SAP UK average Manchester location and Glasgow is compared. Therefore, when an EPC is produced at present for a dwelling in Glasgow its SAP calculated yearly energy and hence EPC rating is actually 15% better than it would be if a more localised weather location was used. Similarly, there is a 15% difference in results between the SAP


of 12

calibrated standard test case with the SAP UK average Manchester location and the London weather: therefore a London dwelling modelled in SAP will be penalised by 15% due to the use of non-location specific weather data. 5.0 Conclusion Results suggest that use of localised weather data can have a noteworthy effect on energy estimation and can play an important role on delivering buildings which are truly fit for purpose. The use of CIBSE TRY/DSY weather files has been shown to provide a wide variation of results, with energy consumption generally increasing with as the geographical latitude of the CIBSE weather location. The spread and variation of the CIBSE weather files has been highlighted: the most accurate weather data to use is that which is closest to the location where the modelled building exists or will exist. The CIBSE weather file spread is good and most major population centres have been well covered but there are large geographical areas of the British Isles where no weather data exists. For delivering buildings which are truly fit for purpose, internal heat gains and building location are becoming more important at the design stage and also once a building is occupied. The research here has highlighted that weather data can have a significant bearing on simulated kWh/year output: a swing of +-15% in calculated energy consumption can be seen based purely upon the external temperature variation simulated in IDEAS from the 14 CIBSE weather locations. 5.1 Discussion and Future Work For the purposes of this study, the focus was the effect upon yearly external temperature profiles to energy estimation of a calibrated dwelling. Future work could be carried out by making use of the solar data available in the CIBSE TRY/DSY weather data, as solar gain can also be an important factor in assessing the energy performance of a building. The effect of rain on U-values of structures could also be assessed using the IDEAS method. This is a large development task but one which can be addressed in IDEAS by the use of dynamically varying U-values, which is another aspect where IDEAS can be used to extend simplified methods such as SAP. A new version of SAP is currently in development, SAP 2012, which will look to take into account regional variations in temperatures for costs but will still make use of a centralised climate location to determine the ratings of dwellings. Once SAP 2012 has been fully published, further work can be conducted by comparing the climactic areas selected by SAP and the climactic areas as defined by the CIBSE TRY/DSY weather location dataset. Furthermore, SAP 2012 results taking into account regional temperatures variations can be contrasted with IDEAS results using regional temperature locations to clarify the outcome of using a monthly data time-step as used in SAP against a minutely as used in IDEAS. SAP 2012 will be used to help assess properties for the Green Deal, the Energy Company Obligation (ECO) and the Renewable Heat Incentive (RHI) [21]. The draft SAP 2012 methodology has now been published and is “anticipated to come into operation in 2013” [22]. The importance of future SAP based developments highlights the need for future work to interrogate the accuracy of SAP methods and to explore any benefits of future SAP variants moving to a more dynamic platform.


of 12

The analysis of the CIBSE TRY/DSY weather locations did raise a number of queries. The distribution spread (Figure 5) highlighted that there is a lack of climactic data for large areas of the British Isles. For example, no weather data exists for areas north of Scotland’s Central Belt. Future work could take weather data from another source, such as Meteonorm, for areas such as Aberdeen and analyse this accordingly. These results could suggest if additional climactic regions should be considered for conclusion in the main CIBSE TRY/DSY dataset. The building design vs. building performance gap will increase the importance of tools used to model buildings in the future; furthermore recent research demonstrates the importance of SAP and its current place in the regulatory framework [23]. This highlights the dwelling design versus dwelling performance gap and the importance of rigour in methods such as SAP. Methods such as IDEAS can be an important part of this discussion and can suggest how methods such as SAP could evolve in the future. References [1] European Parliament. DIRECTIVE 2002/91/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 16 December 2002 on the energy performance of buildings. Official Journal of the European Communities. 2003. [2] Commission E. Directive 2010/31/EU of the European Parliament and of the Council of 19 May 2010 on the energy performance of buildings (recast). Official Journal of the European Union, L. 2010;153:13-35. [3] Uglow CE. Energy use in dwellings: An exercise to investigate the validity of a simple calculation method. Building Services Engineering Research and Technology. 1982;3(1):35. [4] Uglow CE. The calculation of energy use in dwellings. Building Services Engineering Research and Technology. 1981;2(1):1. [5] Murphy GB, Kummert M, Anderson B, Counsell J. A comparison of the UK Standard Assessment Procedure and detailed simulation of solar energy systems for dwellings. Journal of Building Performance Simulation. 2010;4(1):75-90. [6] Tashtoush B, Molhim M, Al-Rousan M. Dynamic model of an HVAC system for control analysis. Energy. 2005;30(10):1729-45. [7] Hudson G, Underwood C. A simple building modelling procedure for MATLAB/SIMULINK. Conference A simple building modelling procedure for MATLAB/SIMULINK, vol. 99. p. 777-83. [8] Cho SH, Hong S, Li S, Zaheeruddin M. An optimal predictive control strategy for radiant floor district heating systems: Simulation and experimental study. Building Services Engineering Research and Technology. 2012. [9] Counsell JM, Khalid YA, Brindley J. CONTROLLABILITY OF BUILDINGS, A multi-input multi-output stability assessment method for buildings with slow acting heating systems. Simulation Modelling Practice and Theory. 2010. [10] Franklin GF, Powell JD, Emami-Naeini A. Feedback control of dynamic systems, 6th edition: Addison-Wesley Reading (Ma) etc., 2010. [11] Tindale A. Third-order lumped-parameter simulation method. Building Services Engineering Research and Technology. 1993;14(3):87. [12] Murphy GB. Inverse Dynamics based Energy Assessment and Simulation. Glasgow: University of Strathclyde, 2012. [13] Griffiths W. Information Paper 10/10. SAP for Beginners; An introductory guide for building professionals. 2010. [14] Mylona A. The use of UKCP09 to produce weather files for building simulation. Building Services Engineering Research and Technology. 2012;33(1):51-62.


of 12

[15] Murphy G, Khalid Y, Counsell J. A Simplified Dynamic Systems Approach for the Energy Rating of Dwellings. Building Simulation 2011. Sydney, Australia 2011. [16] Murphy G, Counsell J. Symbolic Energy Estimation Model with Optimum Start Algorithm Implementation CIBSE Technical Symposium, DeMontfort University, Leicester UK – 6th and 7th September 2011. [17] Beausoleil-Morrison I. Final Report of Annex 42 of the International Energy Agency's Energy Conservation in Buildings and Community Systems Programme: CANMET Energy Technology Centre, Natural Resources Canada, 2008. [18] Meteotest. Meteonorm - Global Meteorological Database version 6.1. 2011. [19] CIBSE. CIBSE Weather Data Packages. CIBSE Knowledge Portal: CIBSE; 2012. [20] BRE. The Government’s Standard Assessment Procedure 2009 for Energy Rating of Dwellings – SAP 2009: BRE, 2012. [21] DECC. Proposed changes to energy assessments of dwellings. 2013. [22] BRE. The Government’s Standard Assessment Procedure for Energy Rating of Dwellings – Draft 2012 edition: BRE, 2013. [23] Zero Carbon Hub. Carbon Compliance for Tomorrow's New Homes. NHBC foundation; 2010.


of 12

Appendix A

Table A1 Building parameters for the Standard Test Case

Parameter Value (Units) Parameter Correlation

zonem 13561 kg Value from SAP

sim 15181 kg Value from SAP

sem 15181 kg Value from SAP

imm 13327 kg IDEAS only, calibration parameter

vm 0.0377 kg/s Value from SAP

zoneC 1129 J/(kg·K) Value from SAP

sC 949.5 J/(kg·K) Value from SAP

imC 1000 J/(kg·K) IDEAS only, calibration parameter

aC 1005 J/(kg·K) Standard Value

wU 1.5 W/(m2·K) Value from SAP

fU 0.7 W/(m2·K) Value from SAP

rU 2.3 W/(m2·K) Value from SAP

imU 2.5 W/(m2·K) IDEAS only, calibration parameter

dU 2.1 W/(m2·K) Value from SAP

sU 1.05 W/(m2·K) Value from SAP

wA 16.9 m2 Value from SAP

fA 44.4 m2 Value from SAP

rA 44.4 m2 Value from SAP

imA 133.3 m2 IDEAS only, calibration parameter

dA 3.8 m2 Value from SAP

sA 81.8 m2 Value from SAP

ih 7.69 W/(m2·K) Standard Value

eh 25 W/(m2·K) Standard Value

wk 0.303 W/(m·K) Value from SAP

wd 0.2375 m Value from SAP


of 12

Appendix B Nomenclature

zonem - Mass of internal air and furniture in environmental zone

sim - Mass of internal structure

sem - Mass of external structure

imm - Internal mass

vm - Infiltration rate

zoneC - Thermal capacity of environmental zone (air & furniture)

sC - Thermal capacity of building structure

wU - U-value of windows

fU - U-value of floor

rU - U-value of roof

dU - U-value of doors

sU - U-value of structure

imU - U-value of internal mass

wA - Surface area of windows

fA - Surface area of floor

rA - Surface area of roof

dA - Surface area of doors

imA - Surface area of internal mass

sA - Surface area of structure

ih - Internal air heat transfer coefficient

eh - External air heat transfer coefficient

wk - Equivalent thermal conductivity of the structure

wd - Thickness of the structure

Benchmarking the effect of weather data upon …...CIBSE Technical Symposium, Liverpool John Moores...

Documents

Transcript of Benchmarking the effect of weather data upon …...CIBSE Technical Symposium, Liverpool John Moores...